A comprehensive security and privacy audit engine for AI coding tools that generates detailed graded reports on telemetry and permissions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-auditor
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-auditor using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agent Auditor?
Agent Auditor is a powerful diagnostic tool designed to provide transparency into the inner workings of AI-powered development environments. As part of the Openclaw Skills ecosystem, it performs a deep-dive analysis of tool source code to identify telemetry, remote control mechanisms, and privacy risks. By evaluating how an AI agent handles your data, this skill ensures that developers can maintain full control over their codebases and intellectual property. It bridges the gap between using powerful AI tools and maintaining strict security standards, providing the clarity needed for professional development environments.
Agent Auditor Use Cases
- Evaluating the privacy impact of a new AI coding assistant before enterprise adoption.
- Detecting hidden telemetry or data exfiltration endpoints in local AI tools.
- Auditing permissions to ensure an AI agent isn't overprivileged in your environment.
- Monitoring for hidden remote control features or killswitches that could impact tool availability.
- Generating security compliance reports for Openclaw Skills and related AI integrations.
How Agent Auditor Works
- The user points the Auditor to the source directory of the target AI coding tool via the CLI.
- The engine performs a deep scan of the source code using advanced pattern matching and static analysis.
- It categorizes findings into specific domains including telemetry, remote control, permissions, and network activity.
- The system calculates a weighted score based on predefined security benchmarks and weighted categories.
- A final report is generated, providing an A-F grade and actionable recommendations for the user to improve their security posture.
Agent Auditor Setup
To get started with this audit engine, you can run it directly via npx to analyze any local tool directory:
npx @claws-shield/cli audit <path-to-tool-source>
Alternatively, for programmatic integration within your Openclaw Skills workflow, you can use the provided script:
node scripts/run-audit.mjs <path-to-tool-source>
Agent Auditor Data Schema & Taxonomy
The audit generates a structured report containing the following data points to help you manage Openclaw Skills safely:
| Data Point | Description |
|---|---|
| Overall Grade | A letter grade from A to F based on the composite score (0-100). |
| Category Scores | Individual performance breakdown across Telemetry, Remote Control, and more. |
| Evidence | Specific source code locations and line numbers where issues were detected. |
| Network Hosts | A list of outbound 1st and 3rd party hostnames identified during the scan. |
| Actionable Items | Specific steps to mitigate identified privacy or security risks. |
Agent Auditor Advanced Features
- Automated detection of undercover modes that strip AI attribution from code or manipulate commit messages.
- Identification of managed settings and accept-or-die dialog mechanisms that limit user choice.
- Mapping of unreleased features hidden behind internal feature flags to track development readiness.
- Comparative baseline analysis against other popular Openclaw Skills and AI tools to rank privacy performance.
- Support for multi-agent auditing to ensure complex, multi-tool workflows remain secure and private.
SKILL.md
Loading
Related Openclaw Skills
McDonald's MCP Integration
A comprehensive integration for the McDonald's MCP platform that enables AI-driven food ordering and loyalty management.
X5 Protocol Skill
A comprehensive CLI tool for executing, debugging, and generating X5 protocol-compliant API requests and cURL commands.
Agent Gateway
A smart multi-model routing proxy for Openclaw Skills that enables seamless switching between major LLM providers with built-in cost optimization and fallbacks.
Lab2Startup Team Selector
A strategic framework designed to analyze and select the optimal leadership model for teacher-student co-creation startup teams.
MIT Phase 6: Funding and Operations
A professional toolkit for calculating capital requirements, designing financing strategies, and establishing scalable operational systems based on the MIT Disciplined Entrepreneurship framework.
Clawtip Pre-Verify
A robust payment utility for Openclaw Skills that handles third-party transaction processing, user token management, and secure wallet access.
Featured*
